Simultaneous Trace Identification and Quantification of Common

Subscriber access provided by ORTA DOGU TEKNIK UNIVERSITESI KUTUPHANESI

Article

Simultaneous trace identification and quantification of common types of microplastics in environmental samples by pyrolysis-gas chromatography-mass spectrometry Marten Fischer, and Barbara M. Scholz-Böttcher Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b06362 • Publication Date (Web): 09 Apr 2017 Downloaded from http://pubs.acs.org on April 17, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Environmental Science & Technology

PET Fiber ~ 1µg

3.9*104 1.7*107

m/z 163

Intensity

1 mm

TIC

Intensity

Page 1 of 31

20 0

20

40 60 Time (min)

40 80

ACS Paragon Plus Environment


1

Simultaneous trace identification and quantification

2

of common types of microplastics in environmental

3

samples by pyrolysis-gas chromatography-mass

4

spectrometry

5

Marten Fischer and Barbara M. Scholz-Böttcher*

6

Institute for Chemistry and Biology of the Marine Environment (ICBM), Carl von Ossietzky

7

University of Oldenburg, P.O. Box 2503, D-26111 Oldenburg, Germany.

8

KEYWORDS microplastic, trace analysis, identification, quantification, simultaneous analysis,

9

Pyrolysis-GCMS, PE, PP, PET, PS, PVC, PC, PA6, PMMA, environmental samples

Page 2 of 31

10 11

ABSTRACT

12

The content of microplastics (MP) in the environment is constantly growing. Since the

13

environmental relevance, particularly bioavailability, rises with decreasing particle size, the

14

knowledge of the MP proportion in habitats and organisms is of gaining importance. The reliable

15

recognition of MP particles is limited and underlies substantial uncertainties. Therefor

16

spectroscopically methods are necessary to ensure the plastic nature of isolated particles,

17

determine the polymer type and obtain particle count related quantitative data. In this study


1

Page 3 of 31


18

Curie-Point pyrolysis-gas chromatography-mass spectrometry combined with thermochemolysis

19

is shown to be an excellent analytical tool to simultaneously identify and optionally quantify MP

20

in environmental samples on a polymer specific mass related trace level. The method is

21

independent of any optical preselection or particle appearance. For this purpose polymer

22

characteristic pyrolysis products and their indicative fragment ions were used to analyze eight

23

common types of plastics. Further aspects of calibration, recoveries, and potential matrix effects

24

are discussed. The method is exemplarily applied on selected fish samples after an enzymatic-

25

chemically pretreatment. This new approach with mass-related results is complementary to

26

established FT-IR and Raman methods providing particle counts of individual polymer particles.

27 28

INTRODUCTION

29

Since the 1950s around 6.46 billion tons of plastics have been produced worldwide (estimated

30

from 1), with 311 million tons in 2014 alone - and the production of MP is still rising.

31

Conservative estimates predict that around 10% of all produced plastics will end up in the

32

oceans 2. Key reasons for this are insufficient waste management and less reflected end

33

consumer habits. Estimations indicate that 2010 alone between 4.8 to 12.7 million tons of land

34

based plastic waste entered the oceans 3.

35

The attention on plastic litter particularly on microplastics (MP) smaller than 5 mm has risen

36

greatly in recent times. The proportion of macro plastics on marine debris has leveled off in

37

recent times 4, that of MP increased continuously mostly due to fragmentation processes of

38

altered plastic litter and additional input of technical micro particles from different sources. In

39

relation to particle numbers it is supposed to be the most prominent plastic fraction today 5.


2


Page 4 of 31

40

World ocean models comparing predicted with measured data show an evident minor presence

41

of measured small microplastics (S-MP) below 1 mm and lead to the assumption of a so far

42

unknown sink of this fraction 6,7.

43

Although the impact of MP on the environment grows with declining particle size 5,8–12, practical

44

specific and selective approaches for qualitative and quantitative analysis of individual plastics in

45

environmental samples on a sub millimeter scale are still lacking.

46

Synthetic polymers are broad in variety in terms of chemical composition and physico-chemical

47

characteristics. Identification of macro- and mesoplastics is possible according to their source

48

and appearance. These features become less distinctive with decreasing particle sizes. Even

49

while using a microscope, an additional, reliable identification method is obligate. For this

50

purpose FT-IR and Raman techniques are the most common used for identification of different

51

polymer types after optical preselection of conspicuous particles. More recently imaging

52

techniques have become popular, which enable an almost automatic scanning of prepared

53

samples for characteristic absorption bands of selected polymers. However, this technique is time

54

consuming, because analysis time increases substantially with decreasing particle size. In all

55

cases particle number related data are generated. The potential of these techniques is reviewed in

56

the literature 13–16.

57

Pyrolysis-gas chromatography in combination with mass spectrometry (Py-GCMS) can be used

58

for reliable identification of isolated plastic particles by analyzing their characteristic thermal

59

degradation products e.g. 17–20 and is commonly applied in the polymer industry. However, this

60

technique has only rarely been used to detect plastics in environmental samples 21–25 but became


3

Page 5 of 31


61

more popular recently 26–28. In these studies particles, suspected to be plastic, are isolated

62

manually and subjected to Py-GCMS.

63

In addition, to identification Py-GCMS could also be applied for quantitative trace analysis of

64

MP on a polymer specific level if the pyrolysis conditions are highly reproducible to generate a

65

consistent composition of pyrolysis products 18,29–31. In environmental samples this was seldom

66

done to date and when attempted it was restricted on a few selected polymers (polyvinyl chloride

67

(PVC) and polystyrene (PS) 22, PS 30 and modified polyamides32. Very recently a promising

68

thermo-gravimetric method for PE determination in environmental samples was introduced.

69

Combined with GCMS this technique bears a great potential for polyethylene (PE) detection in

70

complex samples even with MP concentration as low as 5w% MP so far tested in spiked

71

environmental samples exceed realistic levels in most cases 33.

72

Our intention was to develop an analytical setup, which enables a qualitative and quantitative

73

polymer specific, weight-related trace analysis of MP in environmental samples with a special

74

focus on small MP detracted from optical detection or mechanical particle separation.

75

Accordingly, we demonstrate here Curie point (CP)-Py-GCMS combined with

76

thermochemolysis (a pyrolytic methylation step 34,35) as an equally reliable and practical

77

analytical method for eight relevant common consumer user plastics (PE, polypropylene (PP),

78

PS, polyethylene terephthalate (PET), PVC, poly(methyl methacrylate) (PMMA), polycarbonate

79

(PC), and polyamide 6 (PA6 detected simultaneously within a single GCMS run. Together they

80

represent around 80 % of the actual plastic demand 1 and are prominent in perishables as well as

81

disposable packaging. We show that this approach delivers qualitative information and weight-

82

related at least semi quantitative data for individual polymers on a trace level independent of

83

particle size, shape and a prior optical detection. According to this it is complementary to


4


Page 6 of 31

84

common particle related FT-IR- and Raman techniques for MP analysis. Its successful

85

application in combination with complex environmental matrices is exemplarily demonstrated on

86

fish samples. Their stomachs and gastro-intestinal (GI) tracts contain a broad spectrum of various

87

food related inorganic (e.g. sand, mussel shells) and organic matrix compounds that are

88

representative for those expected in environmental samples. The method can be transferred to

89

any sample type after an adequate MP preconcentration.

90

Concentration ranges to which the method is applicable, detection limits for individual polymers,

91

potential interferences of biological matrices with indicator signals selected and recoveries of

92

polymers spiked into fish stomach samples in trace amounts are presented and critically

93

discussed.

94

EXPERIMENTAL SECTION

95

CP-Pyrolysis-GCMS/Thermochemolysis. CP-Pyrolysis-GCMS measurements were performed

96

with a Curie-Point-Pyrolyzer (CP) Pyromat (GSG Mess- und Analysengeräte, Bruchsal,

97

Germany) in a 590°C CP-pyrolytic target cup (GSG). The Pyrolyzer was attached to an Agilent

98

6890 N gas chromatograph equipped with a DB-5MS-column linked to an Agilent MSD 5793 N

99

mass spectrometer. Additional details on the Py-GCMS conditions are included in the

100

Supporting Information (SI), table S1. Thermochemolysis was performed by adding 10 µl of

101

tetramethyl ammonium hydroxide (TMAH, 25% in water) to the individual samples directly into

102

the CP-pyrolysis targets prior to pyrolysis.

103

Polymer identification. To unequivocally identify single polymers in complex natural samples

104

specific indicator compounds for each polymer are needed. Therefore a database of different

105

polymers (PE, PP, PS, PVC, PA6, PMMA, PET, and PC) was created by measuring polymer


5

Page 7 of 31


106

standards and in some cases additionally polymers of consumer products (for further

107

specifications cf. SI Tab. S2) with direct pyrolysis and thermochemolysis, respectively. The

108

obtained pyrograms were compared with an in-house database and literature data from Tsuge et

109

al. (2011), respectively. The most abundant and/or polymer-specific compounds from TMAH

110

pyrograms (cf. results and discussion) were chosen as indicators for polymer specific qualitative

111

and quantitative analysis (Tab. 2, and SI Fig. S1 - S8, Tab. S3).

112

Calibration. To obtain external calibration curves pyrolysis targets were prepared by adding

113

Al2O3 as inert dilution matrix. Between 0.4 and 1070 µg of polymer standards were weighed

114

directly into the pyrolysis targets using a Cubis® Ultramicro balance MSE2.7S-000-DM

115

(Sartorius, Germany). Finally an Al2O3 cover and in most cases TMAH were added. Polymer

116

standards were measured individually and in mixtures. Individual polymers were calibrated

117

externally by means of preselected indicator compounds by using the mass chromatograms of

118

their respective indicator ions and respective integration results. The bands of confidence and

119

prediction at 95% confidence level each were calculated with Origin 2016G (OriginLab

120

corporation).

121

Fish samples. The applicability of the method to real environmental samples was demonstrated

122

by analyzing fishes caught in 2014. Pelagic and demersal fish samples for recovery studies were

123

caught in the Jade Bay (North Sea, Germany, stow net) and in two locations from the Baltic Sea

124

(Wismar Bay, gillnet, and 35 km north of Rügen Island, demersal trawl). The stomachs (mostly

125

pooled in numbers of five) and gastrointestinal (GI)-tracts used for spiking had samples weights

126

between one and 20 g. All fishes belong to a sample pool of a pilot study on microplastics in

127

fishes; the corresponding data will be published subsequently.


6


Page 8 of 31

128

Sample Clean-up. Sample treatment was adapted and optimized according to existing methods

129

36,37

130

SDS, protease, chitinase and H2O2 treatment to remove biological matrix as much as possible

131

and to preconcentrate potential microplastic content. Subsequently the vacuum dried samples

132

were degreased with a few milliliter of petrol ether (60/80). In case of benthic fish samples a

133

high sand content made an additional density separation with sodium iodide solution (~1.6 gL-1)

134

necessary. All treatments and washing steps were performed in the same crucible, almost the

135

whole time covered with alumina foil to prevent secondary contamination with plastics and

136

fibers from the surrounding air. Parallel to the fish samples procedural blanks were conducted to

137

follow possible contamination pathways and estimate their extent. At the end of treatment the

138

content of crucible was transferred quantitatively to an Anodisc ® filter (0.2 µm) and dried in a

139

glass Petri dish. The 6 mm (∅) filter section containing the sample was stamped out and milled

140

in a small agate bullet mortar, transferred into a pyrolysis target and pyrolyzed after addition of

141

TMAH (25% w/w in water). A detailed clean-up protocol is given in the supplement (SI, S-13).

142

Spiking and recovery experiments. The method was verified on fish samples from the same

143

type already described, spiked with the respective eight polymers at different stages of sample

144

pretreatment (Fig. 2). Around 60 µg (100 µg in case of PE) or less per polymer was spiked, as a

145

warranty of fitness for MP trace analysis. Former analysis data from the same fish set (to be

146

shown elsewhere) ensure that the MP content of the spiked fishes was far below 20 µg for PE, 6

147

µg for PVC as well as PS and 1 µg for PET, PMMA, PA6, and PC, or even lower. PP was not

148

detected in this sample set. Therefore a potential impact of these plastics should be negligible to

149

the general assessment of the method. To investigate possible effects of any remaining matrix

150

(tissue, gastric contents), polymers were added directly to pyrolysis-targets containing the fish

. The samples were enzymatically and chemically digested as a whole via a succession of


7

Page 9 of 31


151

sample. To check the sample digestion procedure polymers were directly spiked to isolated

152

stomachs. The influence of the sample homogenization and transfer step into the targets was

153

tested by adding the polymers to the prepared sample on dry Anodisc® filter.

154

Potential interferences. As potentially interfering compounds which might be present because

155

of incomplete sample digestion or due to secondary contamination pyrograms of chitin, (isolated

156

from samples and copepods), pine wood, wool and cellulose (fiber and tissue) were reviewed for

157

polymer indicator compounds/ions chosen.

158

RESULTS AND DISCUSSION

159

Polymer identification and indicator compound selection. Since respective polymers underlie

160

different decomposition mechanisms e.g. 38,39 their resulting characteristic pyrograms vary

161

basically in terms of generated products, signal number and related intensities. Table 1 gives a

162

synopsis of the pyrolytic behavior of the studied polymers without and with thermochemolysis.

163

The related pyrograms are presented in the supplement (SI Fig. S1 - S8). If pyrolysis is

164

performed after TMAH addition the pyrolytic behavior stayed unaffected for PE, PP, PS and

165

PVC while that of PET, PMMA, and PA6 changed. Since the intention of this study strives for

166

the simultaneous trace detection of the selected polymers, most intensive pyrolysis products are

167

desirable for most sensitive polymer detection. Taking this into consideration, thermochemolytic

168

pyrolysis with TMAH, an online esterification, transesterification, and methylation process 34,35,

169

bears the most benefits.

170

Leading to substantially less and specific reaction products, TMAH addition enhances the

171

detection sensitivity for PET and PC (cf. SI Fig. S7, S8).


8


172

Page 10 of 31

Table 1. Pyrolytic behavior of polymers with and without TMAH addition Polymer

PE

Direct pyrolysis a, c

Thermochemolysis (TMAH)a

Mechanism

(Main-)Signals

Mechanism (Main-)Signals

RCS

Multiple signals of n-alkanes, n- Unaffected

Unaffected

alkenes and n-alkadienes PP

RCS

Multiple signals of methyl alkenes Unaffected

Unaffected

and alkadienes PS

ECS, RCS

Mono-, di-, tri- and tetramers

Unaffected

Unaffected

PVC

Chain

HCl b, benzene, multiple small

Unaffected

Unaffected

stripping

signals of aromatic compounds

PET

CS,

Multiple signals of benzoate and

PBT

secondary

terephthalate derivatives and

reactions

oligomers

PC

Dimethyl TCTM terephthalate

RCS, cross- Bisphenol A 17; triphenyl

Methylated TCTM

linking PMMA Poly(alkyl

phosphine oxide (flame retardant)a

Bisphenol A Methyl methacrylate

Monomer, smaller signals of ECS

dimers and trimer

TCTM

and

alkyl

methacrylates

metacrylate)s PA 6

Methyl

ECS

ε-Caprolactam, traces of nitriles TCTM and cyanopentyl amides

(partial)

ε-Caprolactam,

N-

methyl-ε-caprolactam

173 174 175

RCS = random chain scission; ECS = end chain scission; CS = chain scission TCTM = thermochemolytic transmethylation; a For respective pyrograms in detail see supplement; b according to 38,39; c number of signals depending on polymer concentration

176

Furthermore advantageous is the inclusion of other less used polyalkylene terephthalates like

177

PBT (polybutylene terephthalate) into the measurements due to the generation of the identical


9

Page 11 of 31


178

reaction product, dimethyl terephthalic acid, like PET. This is also appropriate for PMMA and

179

poly(alkyl methacrylate)s leading to methyl methacrylate in varying portions due to pyrolytic

180

transesterification. For PA6, TMAH treatment leads to a partial methylation of ε-caprolactam.

181

The selected indicator compounds for simultaneous detection of the different polymers and

182

related information are summarized in table 2 (cf. also SI Fig. S1 - S8). The determination of

183

most polymers (PE, PP, PVC, PET, PC, PA6 and PMMA) was unambiguous, with regard to their

184

most abundant and selective pyrolysis product(s). For PE n-alkadienes are most indicative

185

pyrolysis products 33. However, their signal intensity decreases drastically at concentrations

186

below 50 µg and they become unreliable for natural samples with low PE concentrations. Here

187

knowing full well that precaution is necessary for forthcoming identification and quantification

188

of PE in environmental samples, the more prominent, but matrix interfered n-alkanes and n-

189

alkenes are chosen for PE determination. This will be discussed and explained later. For PP all

190

three tetramethyl-undecenes and dimethyl heptene, for PET dimethyl terephthalate and for PC

191

dimethyl bisphenol A were selected as indicator compounds. For PVC benzene was chosen as an

192

indicator since hydrochloric acid is inappropriate for GCMS detection and other compounds do

193

not have sufficient sensitivity. PS has two favored indicator compounds, styrene and its trimer

194

(5-hexene-1, 3, 5-triyltribenzene), which differ in specificity and abundance. While the former is

195

very abundant but nonspecific, the opposite is true for the latter. Therefore, styrene is a perfect

196

indicator compound for PS quantification in matrix-free samples such as vehicle paint 30. In

197

natural matrices, however, it may be generated from different precursors i.e. chitin as discussed

198

later. In these cases the less intensive styrene trimer is more reliable since its generation is

199

unequivocally linked to the presence of PS. Spontaneous gas phase polymerization might lead to

200

a loss of pyrolytic generated styrene monomer 40 and influence PS quantification. Even no


10


Page 12 of 31

201

indication for this was observed (cf. SI) it favors the trimer as indicator compound as well. ε-

202

Caprolactam, indicative for PA6, is partially methylated after thermochemolysis. Either signal is

203

very specific for this polymer therefore both are used for identification and the sum of both for

204

quantification.

205

In summary, the selected characteristic indicator products and their respective ions enable a

206

simultaneous detection and identification of the eight selected polymers on a trace level

207

whenever samples are pyrolyzed with TMAH. Although the detection of dimethyl terephthalate

208

indicates the presence of all poly(alkylene terephthalate)s, we use in the following the term PET,

209

the by far main representative, in the following. This applies also to poly(methacrylate)s –

210

containing other substituents than methyl. After thermochemolysis they are partially detected as

211

methyl methacrylate monomer and summarized as “PMMA” disregarding their former molecular

212

weight or substituent. In environmental samples the so-called PMMA signal also includes parts

213

of other methacrylate based polymers as well, but points to their anthropogenic impact in

214

general. Finally, copolymers that contain at least one of the selected plastic types will be

215

determined, too. Thus, the styrene part of, e.g., SAN (styrene acrylonitrile copolymer) will

216

contribute to the determined PS content.

217

Full scan measurements allow the identification of the respective indicator compounds via their

218

mass spectra. Concentration dependent less abundant pyrolysis products enable the identification

219

of the individual polymers, too (cf. Tab. 2).

220

Calibration. The optimal mass range to use Py-GCMS for quantification is polymer dependent.

221

The respective calibration ranges are listed in the SI Tab. S3. Additionally, particle size should

222

be as small as possible to enable optimal pyrolysis and, in case of TMAH use, optimal reaction -

223

surface.


11

Page 13 of 31

224


Table 2. Polymer related pyrogram information Polymer Characteristic decomposition product(s)

M

Indicator ions

MP amount required

(m/z)

(m/z) a

(µg)

S/N

Alkanes (e.g. C20)

2000

282

99, 85

Simultaneous Trace Identification and Quantification of Common

Recommend Documents